Centromeres are critical components of eukaryotic chromosomes, with a key role in ensuring proper segregation in mitosis and meiosis. While the location of the centromere is precisely determined and maintained in most organisms, the basis for centromere specification in many eukaryotic genomes, including the human genome, is obscure and likely involves both epigenetic and sequence-based events. Centromeres represent an evolutionary paradox: despite their essential function in chromosome segregation and the highly conserved nature of many proteins involved in the process of cell division, the underlying genomic sequences are highly variable, both within and between species. In the human genome, centromeres are characterized by large arrays of a tandemly repeated DNA sequence, a satellite. While genetic, genomic and functional studies have demonstrated that a satellite sequences are involved in centromere function in human cells, the sequences are highly heterogeneous and share few features in common with satellite DMAs of non-primate species. Thus, notwithstanding a clear role for epigenetic regulation in specifying centromeric chromatin, our poor understanding of the role of genomic sequences in centromere specification remains a significant gap in current knowledge. The experiments described here have two specific aims: (i) to improve and validate novel human artificial chromosome technology to generate structurally definable, unit-sized human artificial chromosomes that maintain the size and structure of the input vector sequences and can be recovered from human cells for detailed analysis; and (ii) to use human artificial chromosomes to systematically evaluate the role of genomic sequences and their organization in centromere specification in cultured cells, by altering specific sequences within the human a satellite repeat unit, by designing and testing novel multimeric a satellite array configurations, and by substituting in whole or in part other mammalian centromeric satellite sequences from both other primate and the mouse genomes. These experiments will allow us to explore the nature of the genomic code that specifies centromere identity and function despite lack of rigid sequence conservation, as well as provide insights into the genomic and epigenetic mechanisms that contribute to centromere function in human chromosomes. [unreadable] [unreadable] [unreadable]